A modified FCM algorithm for MRI brain image segmentation using both local and non-local spatial constraints

A new segmentation method of cerebral MRI images based on the fuzzy c-means algorithm

“…It is then possible to carry out a regulation according to the similarity degree. We also used the local and nonlocal distance presented by Wang et al [23]. Wang et al modified the initial FCM function by weighting local information and nonlocal information and by redefining the distance between the intensity of a voxel and the centroid of a class.…”

“…Sathya et al [22] used a quadratic regulating term. Wang et al [23], Cai et al [24], Ahmed et al [25], and Bazin and Pham [26] incorporated special information in the regulating term. The limits of the proposed improvements have led us to introduce in our method an approach that can minimize noise sensitivity while taking into account the spatial information of pixels.…”

Improved Bias-corrected Fuzzy C-means Segmentation of Brain MRI Data

“…In addition, the spatially coherent nature of the CLIC criterion function that omits the local gray level relationship makes the estimated membership functions inaccurate, and finally causes misclassifications, especially for the pixels around the boundaries. In order to reduce the noise effect during segmentation, Wang et al [13] introduced a method that incorporates both the local spatial context and the non-local information into the standard FCM algorithm using a dissimilarity index instead of the usual distance metric. Ji et al [14] follow up on Wang et al's approach by using the local spatial context and non-local information to develop the possibilistic fuzzy c-means clustering algorithm PFCM.…”

Harmony search-based cluster initialization for fuzzy c-means segmentation of MR images

“…In this application context, fuzzy clustering has shown tremendous potential as it can naturally cope with such data characteristics. It is therefore unsurprising that the fuzzy c-means algorithm (FCM) [2] has found numerous applications in image segmentation problems and produced very good results [3], [4]. However, the basic FCM algorithm has several limitations and has prompted researchers to investigate various improvisations.…”